INDEX

Explanations

phrases encouraging personal evaluation or decision-making

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

stÃŃ

-0.15

reek

-0.15

åį

-0.14

enu

-0.14

elight

-0.14

andom

-0.14

 Greg

-0.14

miner

-0.14

.rd

-0.14

ÏĢÎ¿Ïį

-0.14

POSITIVE LOGITS

 yourself

0.35

 yourselves

0.28

 themselves

0.28

 Yourself

0.28

 herself

0.25

èĩªå·±

0.23

 ourselves

0.23

 ÑģÐµÐ±Ðµ

0.23

 sobie

0.22

à¹Ģà¸Ńà¸ĩ

0.22

Activations Density 0.054%

phrases encouraging personal evaluation or decision-making

No Comments

No Known Activations